18. Exceptional Human Longevity: The changing pattern of mortality and morbidity among the aged has been exemplified by a consistent pattern of declining disability for nearly two decades accompanied by uneven mortality declines. Both younger and older aged people have experienced improvements in morbidity and mortality. In this context, the most rapid declines in mortality have been observed at exceptionally old ages. While the study of exceptional longevity has focused on mortality registry data and extant demographic models and methods, the study of exceptional healthy longevity requires additional resources in the form of linkage to registries of health service use and surveys which collect health and functional information both across the dimensions of these concepts using widely accepted measurement models and across time using a sampling plan which includes sufficient numbers of exceptionally aged persons for estimation (counts). In addition, national sample surveys corresponding to national registry coverage are most desirable. Among the national population surveys which measure health and function, the National Long Term Care Survey is unique in that it both represents the entire aged population, measures health in detail, and includes large numbers of exceptionally aged persons in the sample design. Indeed, in 1994 and 1999 NLTCS waves, large oversamples of 95+ persons were carried out. One major goal of the project will be to develop information on the persons who eventually attain 95+ in the NLTCS. Health and function histories will be developed over all waves of participation and linked to appropriate registries to identify and understand pathways to exceptional longevity and exceptional healthy longevity. Further, the 1994 and 1999 oversamples will be used to examine recent disability declines at the tail of the survival curve in detail. In particular, it is important to observe that current trends in mortality decline will produce large members of exceptionally aged persons over the next three decades, resulting in a new and distinct population phenomenon. Our specific aims will identify the characteristics and antecedents of this "new" population group as well as the nature of the morbidity/mortality relationship at the tail of the survival curve.